Multi-objective balanced partitioning method for marine sensor network

Existing marine sensor networks acquire data in sea areas divided geographically and store them in their own data centers of the sea area independently. In the case of marine disaster across multiple sea areas, which needs to retrieve data from multiple data centers, the current network structure has a serious impact on real-time decision making. To efficiently carry out data layout and speed up the data retrieval, in this study, the marine sensor network is abstracted as a graph, all the sensors are considered as the vertexes of the graph, and the degree of correlation between the sensors computed with the previous disasters data is taken as the edge. A multi-objective optimization algorithm (NSGA-II) is used to partition the abstract graph into multiple regions and store them in a cloud computing platform. The NSGA-II maximizes the correlation of sensor within the region, minimizes the correlation of inter-region, achieves a balanced size of the regions and minimizes communication time of inter-region. The experimental results show that the proposed method can achieve the optimal layout for distributed storage in the process of disaster data retrieval in China Sea area, and effectively shorten the data retrieval time, providing fast and efficient data access service for marine disasters.